So what do you think will happen to the animation industry in 20-30 years once algorithm/A...

So what do you think will happen to the animation industry in 20-30 years once algorithm/A.I can churn out hundreds of drawings per second that may take a human artists days to do? Or create a 3D model and animate it in a fraction of the time it takes a team of animators to do so?

Attached: artificial-intelligence-1.jpg (580x437, 56K)

Other urls found in this thread:

techcrunch.com/2018/01/18/microsofts-new-drawing-bot-is-an-a-i-artist/
en.wikipedia.org/wiki/Generative_adversarial_network
twitter.com/SFWRedditVideos

In 20-30 years, it won't look much better
You'll still be able to tell that it's heavily computer generated

What you should be looking out for is a paperman style renaissance when people realize the style is stunning and beautiful in it's own way

There's only so much you can do with 3D animation, only so realistic and fluid you can make something without breaking the model down and rearranging it frame by frame, and after a while it gets stale

>In 20-30 years, it won't look much better
>he actually believes this

Attached: 17B93399-E42E-40E2-922B-45325B453B51.jpg (374x250, 24K)

Mate, computers today are already drawing photorealistic pictures.

techcrunch.com/2018/01/18/microsofts-new-drawing-bot-is-an-a-i-artist/

Attached: drawing-bot-feature-3.jpg (800x461, 83K)

Meh, I hope it gets better. It just looks like a shitty filter.

Having a pc that draws thousand frames would be neat.

Mate something is wrong with your eyes because if I see the final picture, I would not be able to tell the difference between it and a actual picture of a real bird. And pls, what we're doing with A.I now, researchers thought was 20 years away. It's actually scary just how fast the field is advancing. The picture would have been considered impossible even for a supercomputer just 3 years ago.

Art started off being perceived as another form of craftsmanship that deserved no more recognition than the labor or a carpenter or a shoemaker.
Then it became an intellectual discipline, honored and respected by kings and peasants alike.
Now it's headed towards becoming something computers do.
We had a good run; this is just one more step towards humanity becoming souless.

I recall seeing John Lasseter talking about his early computer animation "The Adventures of André & Wally B." (1984). After the screening, he says people came up to him to ask what program he used to make it funny.

Computers may be able to do the gruntwork of applying filters and making calculations and rendering things photorealistically, but they'll still be missing something only a human can bring.

No company will care about the human component. They'll just have the computer do most of the work, add in a few basic formulas and jokes for the plot and start churning them out. When quantity over quantity becomes a viable option they will take it without a second thought.

>Mate something is wrong with your eyes because if I see the final picture, I would not be able to tell the difference between it and a actual picture of a real bird.

It must be because I am an artfag, or used to photoshop filters.

Also, this is not much of a surprise for me.

> implying tweening isn't the definition of soulless torment
AI animation will actually free people and allow more resources to be allocated into actually creative parts of the craft, like design and storytelling.

And the worse part is that it's probably going to be in 3D, just imagine just clicking a button and randomising thousands of background characters, all fully modeled, textured and rigged and having the A.I do the basic animation like walk cycles. Even the main characters can probably have their lip animation and walking/running animation done in a single click. All for no cost and in seconds. 3D shows are already so prominent due to their low cost and this is just going to make it worse.

Once that becomes the norm I wonder what people will think of human made animation.
My bet is on a division between the people who find it sloppy and primitive and another group who will miss those days and be called nostalgiafags.

Attached: 1515342790151.jpg (600x623, 57K)

On the bright side, once machine intelligence gets that powerful, actual human level strong A.I won't be that far away. So all you need to do is hang for for another 10-20 years, hope that that they don't want to kill us all and they could be churning out better then human level animation and story millions of times faster then in all of human history.

> in the near future, the machines do everything
> human existece is reduced to watching machine-made cartoons 24/7 and shitposting about it on Sup Forums
> the machines draw their power from people solving google capcha
Truly a perfect symbiosis.

The simpsons will finally end in 2038

So what happens when A.I start being able to actually draw 2D pictures, like they're starting to be able to do right now?

The distinction between "live action" and "motion capture" animation will fade away.
There are people who can ACT, who can "become" the character the script and director call for.
And there are people who are handsome/gorgeous, but who have all the "depth" of a plank of wood.
I'm sure you've seen both types on-screen and can name names.

GREAT actors are both. Or they play roles which don't demand "matinee idol" looks.

With CGI, "appearance" and "talent" will be separated. You don't expect a good writer to necessarily be able to draw, do you? (Yes, there are a few who can do both. But how many?)

I expect this "division of labor" to result in better movies.

Also, movies have become formulaic and predictable because it takes $$$$ to make a movie. Studios and banks are reluctant to invest in innovation. If you're not beholden to them and can build a city and have thousands of extras within the budget of an "indie" production -- I don't see that as a bad thing.
Movies will become more like books -- the creation of an individual author (and maybe an editor.)

This doesn't sound too bad

pc has made cartoons worse

"pc can make everything better" this anglo mentality destroyed traditional 2d ones and they cant stop it

We could possibly have another explosion of user-created content akin to the Newgrounds boom, since people could actually work full-time and still have time to animate.

Just a reminder that graphene transistors could mean computers thousands of times faster speeds then traditional silicone transistors and use hundreds of times less power. A lot of the A.I advances this days are mostly due to the software, machine learning neural networks be crazy, but combine that with a new revolution in transistors and this shit is going sooner then you think.

Same thing that happened to music now that's it's trivial to produce songs in your garage. The mainstream stuff will suffer immensely as companies exploit it, yet the 'indie' stuff gets way better as people actually have more resources available.

Starry night is van Gogh's expression of the way he sees the stars, as more than just tiny pale dots in the sky.
Can a computer replicate that? The act of sharing your emotions and thoughts through the medium of art?

Yes, seeing as this programs/algorithm will be used by humans to convey their emotions and thoughts, it's just a lot easier then drawing all those thousand of frames of animations themselves. And pls, do you think that normies can tell the difference or even care? Or the retards that visit that place?

its obvious you havent ever made anything with a pc
like sjws in media, typical hyperbolic american journos

What I can make with a computer today, entire cities couldn't a hundred years ago. And pls /po/tard, what part of a modern day life has heart and soul in it anymore? Most people use and eat mass produced cheap plastic products, listen to mass produced focus tested music and watch mass produced focus tested movies. Or do you think that shit like the emoji movie or minions are works of art? And yet they continue to make massive profits anyway. If there is software that allows movies like that to be made with half the cost and 1/10th of the time, who wouldn't jump at it?

gibberish
what the fuck are you even talking about

learn how to use layers in ps, how to use a step sequencer in a dtm software before bullshitting

How far away are these graphene transistors?
And how far away are the versions affordable enough for regular consumers?

>using some cutting edge image-generating computer algorithm
>people will spend shit loads of money on this but won't do 2D because "it's expensive"

Attached: 1515552422476.jpg (250x241, 7K)

And again, that's something which would been impossible to do hundreds of years ago. But someone can easily do so today. And the easier it is, the less passionate you need to be. Stop kidding yourself, people don't care about passion or heart or soul, they just need something entertaining.
That you need to ask the researchers, who probably have no idea themselves. But still, we know it's possible and that's enough. Tell someone 30 years ago that we would be able to fit billions of nano scale transistors onto a chip the size of your palm and that we produce billions of this CPUs and at cheap enough prices that almost everyone can afford it and they would have called you insane.

the r&d is expensive, but once it's there you'll be able to liscence that shit for the price of 3 full time animators.

>Assuming that they don't jack the price up to recoup their R&D cost.

These aren't real people
en.wikipedia.org/wiki/Generative_adversarial_network
this honestly, people are still very much going to be part of the creation process, its the tools that will make self-expression easier.

Attached: ma18-iangoodfellow3a.png (421x432, 134K)

It has no style, no liveliness.
Once a computer can actually create an appealing image with a sense of style then and only then will I cede to this potentially being a problem.

> assuming chinks won't steal the tech and sell it 10 times cheaper cause they have no R&D investments to recoup

Those pictures were made by a computer... OP's pic, also made by a computer. And people will still be involved in the creation process, it could be as easy as drawing a few sample pictures/character designs and asking the computer to copy them/the style you use.

>Computers haven't massively advanced in the last 20 years.

There's a reason why the OP said 20-30 years in the future you retard

I don't think a computer made another
animation would look the same. Animation suudios like trigger when they are animating a crazy movement hey over exaggerate it for a weird effect. Computers can't just create a smear frame, or make animation that looks decent. I feel like utilizing a 3D model for 2d would look so bloody stiff it wouldn't work

Even if they just made backgrounds, or just did the colouring or never touched 2D and just focused on 3D and CGI, it would still have a massive impact.

Much easier, no matter where you look or how you think this kind of software is going to develop.

But who knows where it could go? People used to think that computer drawing photorealistic pictures, being able to recognize images almost as well as a human and beating humans at Go would be in 2040 and beyond. And look where we're now. What if they do manage to algorithms that could animate action scenes well? Or do so in such a way that a human only needs to "touch up" on the software's work, reducing the work and people needed by a tenth-fold?

If it's possible it'll bring down costs in America even further, but they already don't do their own animation so artistically it won't matter. As for anime I think the complexity of the drawings and animation and the very deliberate selection of key frames will make it unviable. Maybe it will work for in-betweening, but even then I doubt it.

Technology will not and cannot advance more than it already has. We have reached the peak of what’s physically possible.

>t. a fucking idiot

Wot?

Attached: 3A039DD1-0555-4A52-B2A5-C89D612673B2.gif (500x281, 1.28M)

Nonsense.
Computers already do "squash and stretch" and deliberate blurring.
Even my out-of-date software lets you specify that a bouncing ball has the properties of iron or the properties of jello. Choose the latter and lower the gravity so you can follow the action and it's everything Tex Avery could have asked for.

We're teaching computers how "real" physics works. Just as easy to program "unreal physics". One subroutine codes for Isaac Newton (F=gMm/r^2) and another codes for Chuck Jones (Gravity only operates once characters double-take and notice there's no floor beneath them)

Making this work for complicated drawings and animations is a different story.

We're talking about 2D you retarded nigger kike

Yeah but rotating that model and taking snapshots for every key frame and flattening itou to give the illusion of 2d might as well just be making a stop motion and running the pictures through illustrator. It'll take a couple years before we even learn to mske that look viable. Let alone having a computer to calculate the best succession of images to give the illusion of weight.

Attached: IMG_3303.png (1136x640, 485K)

>Technology will not and cannot advance more than it already has. We have reached the peak of what’s physically possible.
This is much more true that you realize, moron.

Added with the fact that techincally speaking, character models are visually perfect, whereas 2d animation is not. It's that imperfection that gives 2d its signature feel. Even the best animators like Richard Williams and James Baxter a their best animation looking 3D still was imperfect and I don't think a computer can replicate it.

Attached: IMG_3251.jpg (1032x814, 208K)

Animation as never about having the most realistic drawings, (unless you want rotoscope) it's about using the 12 principals to give the most "realistic" looking movement. It was only over the years that drawing became more complicated while juggling the 12 principals.

At least artists won't be the only group whose skillset becomes obsolete.

I thin there will be a split between the new style and the "old" style. I feel like 2d will make a comeback in a hipster sort of way and will be able to co exist together with everyone else preferring the latter.

I'm surprised that no one has mentioned that there are a couple of songs that were written by A.I. It's entirely in the realm of possibility that computers could end up doing a whole show if you give them a formula to follow for the plot.

When will tech fags learn that certain things cannot be completely replicated by a computer. There's no mathematical reasoning for making characters look "alive" and animated. There will always be a need for human input to make sure that the shit these computers spit out actually looks lively. Can a computer shit out thousands of images in succession? Sure. Is it something people would actually like and find charming? That's another story.

If you actually read the fucking thread, no one is actually saying that. Of course there needs to be human input, but's the industry is going to undergo a massive change if improved algorithms and software allow a single person to be as productive as a dozen animators today.

In america they just story board the key scenes and sent it to Korea for animation, just replace Korea with "my desktop ", of course it's gonna need tons of editing and the more complicated scenes are still gonna need the "human touch" but as long as it reduces the cost and time involved, what company isn't going to eat this shit up?

And that's not mentioning 3D, where computers could really do 95% of the work.

Just think of just asking your computer to make a thousands variations of a forest drawn in the show's art style, all you have to do is pick out the nicest looking one, add in some of your personal touch and there, you just saved yourself a few hours of work. That's a massive amount of work hours that could be saved just from making the computer draw the background for you. And for every big action scene, there's a dozen and one scene where the characters are just sitting/standing around and talking, with minimal movement, which I'm sure could also be animated without much trouble by a computer eventually. Again, that's a massive amount of time saved.

Korea is basically an assistant that inbetweens frames and (digitally) inks the keys. An AI that can predict exactly where the inbetween should go and draws it perfectly on model, is something I want to see. Although for more specialized sakuga, the animator should do that himself.

BS, idiot

in fucking 1985, we didn't think computers needed more than Xkb of fucking ram!

it WILL get better, asshole.